Motion Generation for General Robot-Object Interaction

(Principal Investigator, 5-year grant)

• Period: 2025 – Present
• Funding: Sejong Science Fellowship, National Research Foundation of Korea (NRF)
• Key Contributions:
  • Development of simultaneous hand motion and object tracking technology.
  • Construction of a digital-twin simulation environment and dataset expansion.
  • Development of robot motion generation technology based on real-virtual data co-training.
  • Collection of daily environment datasets and application to real-world robotic tasks.

Humanoid Bimanual Teleoperation System for Precision Manipulation

• Period: 2025 – 2026
• Funding: Robot Center, Samsung Research
• Key Contributions:
  • Development of a teleoperation glove combining haptic feedback and visual–inertial hand tracking.
  • Design of a low-latency, high-precision teleoperation interface for dual-arm/dual-hand humanoid robots.
  • System deployment on bimanual robots with real-time precision-task execution and validation.

Photo-Realistic/Physically-Precise Simulator and Large-Scale Multimodal Datasets for Robot Foundation Model

• Period: 2025 – 2026
• Funding: Ministry of Trade, Industry and Energy (MOTIE)
• Key Contributions:
  • High-precision, photo-realistic physics simulation for contact-rich robot behaviors.
  • Provision of accurate physical signals and real-world–quality visual outputs.
  • Sim-to-Real application on humanoids using Robot Foundation Model–based methods to validate simulators and datasets.

High-Payload Multi-Degree-of-Freedom Aerial Manipulation Robotic Platform

• Period: 2024 – Present
• Funding: Ministry of Trade, Industry and Energy (MOTIE)
• Key Contributions:
  • Development of a bimanual finger-based interface for teleoperation of aerial manipulators.
  • Sensor fusion and SLAM for real-time robot state estimation and 3D digital-twin generation.
  • Establishment of a digital-twin based simulation for physically-accurate robot AI learning.

AI Simulation Platform for Smart Robots

• Period: 2022 – 2024
• Location: Global AI Center, Samsung Research
• Key Contributions:
  • Simulation of real-time robot sensor data (Vision, IMU, LIDAR) and actuator data (BLDC, Servo) in smart home environments.
  • Development of control strategies via AI algorithms (Reinforcement Learning, Transformer, Motion Planning).
  • Contributed to the release of Samsung robots, including robotic vacuum cleaners and Bot Handy.

Development of Large Multimodal Model "Samsung Gauss"

• Period: 2023 – 2024
• Location: Global AI Center, Samsung Research
• Key Contributions:
  • Participated in the development of Samsung Gauss, a proprietary large multimodal foundation model.
  • Contributed to the design and training of vision-language architectures for cross-modal understanding.
  • Explored the integration of Large Multimodal Models (LMMs) with embodied AI to enhance robotic perception and decision-making.
  • Optimized model performance for real-world applications within the Samsung product ecosystem.